MolecularBioSystems
This article was published as part of the
2009 Emerging Investigators IssueHighlighting the work of outstanding young scientists at the
chemical- and systems-biology interfaces
Please take a look at the full table of contents to access the other papers in this issue.
Screening of a branched peptide library with HIV-1 TAR RNAwzDavid I. Bryson,
aWenyu Zhang,
aW. Keith Ray
band Webster L. Santos*
a
Received 2nd March 2009, Accepted 11th June 2009
First published as an Advance Article on the web 7th July 2009
DOI: 10.1039/b904304g
The recognition that RNA is more than just an intermediate in the information transfer from
genetic code to fully functional protein has placed it at the forefront of chemical research. RNA is
important because of its vital role in regulating transcription, translation, splicing, replication and
catalysis. Consequently, molecules that can bind to RNA and control its function have potential
as powerful tools in biology and medicine. Herein, we report the discovery of HIV-1 TAR
RNA-selective ligands using an on-bead screening of a library of 4096 branched peptides.
Introduction
The recognition that RNA is more than just an intermediate in
the information transfer from genetic code to fully functional
protein has placed it at the forefront of chemical research.
RNA is important because of its vital role in regulating
transcription, translation, splicing, replication and catalysis.1
Consequently, molecules that can bind to RNA and control its
function have potential as powerful tools in biology and
medicine.
In principle, the tertiary structure of RNA can provide a
scaffold in which molecules can dock and form the necessary
intermolecular interactions to become highly selective. The
formation of unique three dimensional architecture, often
generated through formation of local structures such as
bulges, stem-loops, pseudoknots and turns, differentiates
RNA from DNA and allows it to become a therapeutic
target.2 In fact, antibacterial drugs (aminoglycoside, macro-
lide, oxazolidinone and tetracycline) have cemented the
drug-feasibility of RNA.3 These molecules exert their
antibacterial activity by binding to specific regions of rRNA.
However, it is clear that targeting generic RNA with small
molecules is difficult and requires further studies.4 Still, the
difficult task is designing ligands independent of the canonical
Watson–Crick base-pairing on RNA primary structure. Thus,
we became interested in the widely recognized problem of
developing RNA-selective molecules.
Many important functions of RNA result from specific
proteins binding to complex tertiary structures of RNA. One
such example is evident in the HIV-1 transactivation response
element region (TAR) RNA, a conserved 59-nucleotide stem
loop located at the 50 end of all nascent transcribed HIV-1
mRNA.5 The TAR structure is comprised of a highly
conserved hexanucleotide loop and a three-nucleotide bulge
flanked by two double stranded stems (Fig. 1). Binding of the
transcriptional activator protein Tat and the cyclin T1–cdk1
kinase complex promotes efficient transcriptional elongation.
Tat is an 86 amino acid protein that contains an arginine rich
motif (ARM), which specifically interacts with the TAR
tri-nucleotide bulge (U23, C24, and U25). Disruption of
Tat–TAR interaction results in the blockade of viral
replication and thus represents a viable strategy in developing
new anti-HIV therapeutics. Several ligands (aminoglycosides,6
argininamide,7 peptides,8 peptidomimetics,9 small molecules,10
and others11) are continuously being developed to inhibit
Tat–TAR interaction (Fig. 1).
Owing to the complexity, the lack of designable ligands and
a growing interest in developing new molecules that can
selectively interact with RNA, we aimed to develop a strategy
that allows for the rapid screening of a short branched peptide
library and determine their capacity as possible ligands for
HIV-1 TAR RNA (Fig. 2). Branched peptides have found
extensive use in biological systems including synthetic peptide
vaccines (multiple antigen peptides), drug delivery vehicles and
therapeutics for a variety of disease states.12 To date, the
potential of using branched peptides for the selective targeting
of structured RNA targets has not been realized. Because
multivalency is often described to increase the affinity of a
ligand to a particular receptor,13 we wanted to determine
whether branched molecules are capable of forming
multivalent interactions with RNA and whether presentation
Fig. 1 Sequence and secondary structure of HIV-1 TAR RNA (A),
small molecule ligands (neomycin B (B) and argininamide (C)) and
11-mer Tat-derived peptide (D) that inhibit Tat–TAR interaction.
a Virginia Tech, Department of Chemistry, Blacksburg, VA 24061,USA. E-mail: [email protected]; Tel: +1 540-231-5742
bVirginia Tech, Department of Biochemistry, Blacksburg, VA 24061,USA
w This article is part of the 2009 Molecular BioSystems ‘EmergingInvestigators’ issue: highlighting the work of outstanding youngscientists at the chemical- and systems-biology interfaces.z Electronic supplementary information (ESI) available: Experimentaldetails including the synthesis and characterization of branchedpeptides; protocols for on-bead screening and fluorescence polariza-tion measurements. See DOI: 10.1039/b904304g
1070 | Mol. BioSyst., 2009, 5, 1070–1073 This journal is �c The Royal Society of Chemistry 2009
PAPER www.rsc.org/molecularbiosystems | Molecular BioSystems
of structurally diverse functional groups can enhance selectivity
with the target RNA. Branched peptides are a good starting
point because naturally occurring amino acids contain a wide
array of substituents that can display unique molecular
architectures, and their synthetic accessibility is straight-
forward. An attractive feature of this strategy is the synthesis
of the library on solid support, allowing for rapid generation
of a large number of peptides using the split and pool
technique, and subsequent on-bead screening against
TAR.8,14,15 We designed our initial branched peptide library
with biased parameters that enhance binding and selectivity to
RNA (Fig. 2). These interactions include electrostatics
(positively charged residues interacting with the negatively
charged phosphate), p–p interactions and hydrogen bonding.
In principle, the diversity of possible peptide structures can be
immense because of the commercial availability of unnatural
amino acid monomers.
Results and discussion
The branched peptide library was prepared such that each
variable position contained one of four possible amino acid
monomers (Fig. 2). Branches A1–A3 had an identical sequence
in order to simplify the de novo sequencing by MALDI/mass
spectrometry (MS) and were branched through the a- and
e-amino groups on lysine. Using orthogonal protecting
groups, 4096 compounds were generated by the split and pool
technique in a few peptide coupling steps (ESIz). It was
important that the peptides were not released from the beads
after deprotection of the side chain residues because the
on-bead screening required resin-tethered probes. Thus,
peptides were attached to TentaGel NH2 beads via a photo-
cleavable linker (3-amino-3-(2-nitrophenyl)propionic acid,
ANP),16 which allowed the efficient release of hit molecules
by UV-irradiation, post screening (Scheme 1).
Previous high-throughput screening methods suggested that
the on-bead screening method can be adapted with our
branched peptide library.8,14 Following this protocol, we
transferred three copies of the library in an Eppendorff tube
and washed them with water (5�) and buffer (3�) containing50 mM Tris HCl, 20 mM KCl, and 0.1% Triton X-100 at
pH 7.4.17 In order to address the issue of selectivity at the
outset, we decided to preclude the beads containing
promiscuous RNA binders. A simple process that can increase
assay selectivity involves the addition of competitor RNA. In
this case, unlabeled a-synuclein mRNA, a 340-nucleotide long
RNA, was added in the incubation mixture. We hypothesized
that longer RNA sequences are likely to bind nonspecifically
to our library, and abolishing promiscuous binders can
decrease the potential for off-target binding. Thus, we
performed a preliminary incubation of a small subset of beads
(B2000) in BSA (1 mg ml�1) and 500 nM unlabeled
a-synuclein mRNA at 4 1C for 3 hours and washed (3�) thebeads to remove unbound BSA and RNA in solution. Finally,
TAR RNA screening was effected by incubating the
29-nucleotide TAR RNA-labeled with DY547 fluorescent
tag on the 50-end at 3.4 nM concentration for 3 hours at
4 1C. The solution was then filtered, washed successively with
buffer, placed in a 96-well plate and analyzed using confocal
microscopy. It was evident from the initial experiment that
the stringency of the assay was inadequate because out of
B2000 beads, six beads were sufficiently fluorescent to
isolate them.
The results of the preliminary incubation studies suggested
that a more stringent protocol is necessary to decrease the
number of hit compounds that will have to be analyzed. To
achieve this goal, the second incubation time was reduced to
1 hour. Following this protocol for the remaining library
members, 10 additional beads were selected. Fig. 3 shows
three representative fluorescent beads visualized under laser
irradiation (Fig. 3A) and transmitted light with laser
irradiation (Fig. 3B). These beads were washed with buffer
and organic solvents to remove the bound RNA until
fluorescence was no longer visible under the confocal micro-
scope. As expected, rescreening of the same beads with
DY547-TAR RNA under a more stringent condition
confirmed RNA binding (ESIz). Finally, labeled RNA was
removed with copious buffer and organic solvent washes,
which was followed by photolytic release of branched peptides
from the resin by exposure to UV irradiation (365 nm,
handheld UV lamp, 4 W) for 1 hour.
The identities of the peptide hits were determined by de novo
sequencing of the resulting solution by MALDI/MS (Table 1).
Sequence homology analysis of the hit peptides revealed a
Fig. 2 Combinatorial library of 4096 branched peptides synthesized
by split and pool technique. Peptides are attached to TentaGel resin
via a photocleavable linker, and the branching unit used was lysine.
Scheme 1 On-bead screening of branched peptide library against
50-DY547-labeled TAR RNA and hit identification using MALDI/MS.
Fig. 3 Visualization of hit beads under confocal microscope (A) and
transmitted light with laser irradiation (B); arrows point to additional
beads. Sequence alignment of 16 hits compounds generated a LOGO
shown in (C).
This journal is �c The Royal Society of Chemistry 2009 Mol. BioSyst., 2009, 5, 1070–1073 | 1071
LOGO shown in Fig. 3C.18 Of the 16 branched peptides,
twelve contained Arg–Arg on the N-terminus—an unsurprising
outcome given that RNA binding can be mediated by electro-
static interactions. This result suggests that a positive charge is
favored both in the A1 and A2 positions. While no
overwhelming preference for a specific amino acid was
observed at any other positions, interestingly, arginine was
not the preferred amino acid in position A5. Because the
propensity for RNA binding is expected to increase in the
presence of an electrostatic interaction, albeit nonselectively,
the preference for tyrosine in position A5 is noteworthy. It
appears that a favorable electrostatic force can be adequately
replaced with potential hydrogen bonding and hydrophobic
interactions. Consistent with this observation, hydrophobic
residues, Phe and Leu, dominate position A6 when possible
amino acids such as Ser and Asp that can hydrogen bond are
present. It was also gratifying to find that one of the beads
collected from the library screening, BP16, contained the
sequence with the highest amino acid frequency as indicated
in the LOGO illustration in Fig. 3C.
To evaluate whether the branched peptide hits bound TAR
RNA, biophysical characterization was undertaken. We
selected BP15, BP16 and BP17 ([RRL]2*HRF) for analysis.
Validation of the activities of these compounds was
paramount because false positives were possible, as in any
high-throughput screening assay. BP15 was randomly selected
from the hit peptides, and although BP17 was not a sequence
derived from the library screening, we hypothesized that this
sequence should have good binding affinity for HIV-1 TAR
based on the consensus sequence analysis. For this purpose,
the dissociation constants for these compounds were
determined using fluorescence polarization (FP). This
technique required the resynthesis of select peptides with FITC
on one of the N-terminal branches, which was accomplished
using orthogonally protected lysine (Fmoc-Lys(ivDde)-OH) as
the branching unit (Fig. 4). The resulting branched peptides
contained a N-terminal acetyl group on one branch. After
HPLC purification and characterization by MALDI/MS, FP
experiments were performed in triplicate for each peptide
(Fig. 5). In each case, 0.5 nM FITC-peptide was incubated
with increasing concentration of TAR RNA (up to 100 mM).
The results reveal that BP15 has a Kd of 27 mMwhile BP16–17
have affinity for TAR RNA in the mM range (Fig. 5). Because
the FP technique required the use of high concentration of
RNA to reach full saturation, in this case 4100 mM, we were
not able to determine the accurate dissociation constants for
BP16–17. Gratifyingly, when a mutant version of TAR
(U24 - C24, ESIz) was titrated against BP15–17, it is clear
that their binding affinity is severely decreased, suggesting that
these peptides are selective towards the TAR RNA used in the
screening. As a control, we also synthesized a linear peptide
version of BP15, T15, to investigate whether branching has an
effect on the binding affinity towards TAR RNA. T15 has a
sequence of RRAGVRD, where the branching unit Lys was
replaced with Gly. The data indicate a marked loss of binding
affinity (410 fold) suggesting that the additional functional
groups present on the branch aid in anchoring to TAR RNA.
Table 1 Sequence of branched peptides BP1–BP16 derived fromMALDI/MS
Sequence Sequence
BP1 (RRL)2*WYL BP9 (RRA)2*NYFBP2 (YRA)2*HRF BP10 (DNL)2*HYFBP3 (HRW)2*WAS BP11 (RRA)2*VYFBP4 (RRW)2*HAL BP12 (RRW)2*HASBP5 (RRL)2*NRF BP13 (RRY)2*NQDBP6 (RRY)2*VRL BP14 (RRY)2*VQLBP7 (RRW)2*HYD BP15 (RRA)2*VRDBP8 (YRL)2*WRL BP16 (RRL)2*HYL
Fig. 4 Synthesis of FITC-labeled peptides. SPPS, solid-phase peptide synthesis; Ac, acetyl; FITC, fluorescein isothiocyanate.
Fig. 5 Fluorescence polarization binding curves of branched peptides
with HIV-1 TARRNA and mutant (U24- C24) TAR RNA. T15 is a
linear peptide (RRAGVRD) version of BP15. Each experiment was
done in triplicate.
1072 | Mol. BioSyst., 2009, 5, 1070–1073 This journal is �c The Royal Society of Chemistry 2009
Conclusion
In conclusion, we discovered selective binders for HIV-1 TAR
RNA using an on-bead screening of a focused library of
branched peptides. The use of branched structures with diverse
functional groups and capacity to form multivalent inter-
actions is important in defining new molecular entities that
can be selective for the tertiary structure of target RNA. In
addition to synthesizing a more complex branched peptide
library, future work in our laboratory will focus on extensive
characterization including mapping of TAR–branched peptide
interactions, cell permeability and in vitro inhibition of
Tat–TAR RNA interactions.
Acknowledgements
We thank Prof. Richard Helm at VA Tech mass spectrometry
incubator and Jason Crumpton for help with sequencing
branched peptides. We also thank Dr Phillippe Bissel for
synthesizing Fmoc-ANP and Dr Kristi DeCourcy at the Fralin
Life Science Institute for assistance with confocal microscopy.
Support for this project was provided by the Department of
Chemistry at Virginia Tech. This material is based upon work
supported in part by the Macromolecular Interfaces with Life
Sciences (MILES) Integrative Graduate Education and
Research Traineeship (IGERT) of the National Science
Foundation under Agreement No. DGE-0333378.
References
1 D. E. Draper, Annu. Rev. Biochem., 1995, 64, 593–620.2 R. Schroeder, A. Barta and K. Semrad, Nat. Rev.: Mol. Cell Biol.,2004, 5, 908–919.
3 T. Hermann, Curr. Opin. Struct. Biol., 2005, 15, 355–366.4 J. R. Thomas and P. J. Hergenrother, Chem. Rev., 2008, 108,1171–1224.
5 F. Aboul-ela, J. Karn and G. Varani, J. Mol. Biol., 1995, 253,313–332; T. M. Rana and K. T. Jeang, Arch. Biochem. Biophys.,1999, 365, 175–185.
6 H. Y. Mei, D. P. Mack, A. A. Galan, N. S. Halim, A. Heldsinger,J. A. Loo, D. W. Moreland, K. A. Sannes-Lowery, L. Sharmeen,H. N. Truong and A. W. Czarnik, Bioorg. Med. Chem., 1997, 5,1173–1184.
7 J. Tao and A. D. Frankel, Proc. Natl. Acad. Sci. U. S. A., 1992, 89,2723–2726.
8 S. Hwang, N. Tamilarasu, K. Ryan, I. Huq, S. Richter, W. C. Stilland T. M. Rana, Proc. Natl. Acad. Sci. U. S. A., 1999, 96,12997–13002.
9 Z. Athanassiou, K. Patora, R. L. Dias, K. Moehle, J. A. Robinsonand G. Varani, Biochemistry, 2007, 46, 741–751.
10 B. Davis, M. Afshar, G. Varani, A. I. Murchie, J. Karn,G. Lentzen, M. Drysdale, J. Bower, A. J. Potter, I. D. Starkey,T. Swarbrick and F. Aboul-ela, J. Mol. Biol., 2004, 336, 343–356;R. Pang, C. Zhang, D. Yuan and M. Yang, Bioorg. Med. Chem.,2008, 16, 8178–8186.
11 M. Froeyen and P. Herdewijn, Curr. Top. Med. Chem., 2002, 2,1123–1145.
12 J. P. Tam, Proc. Natl. Acad. Sci. U. S. A., 1988, 85, 5409–5413;D. Singh, R. Kiarash, K. Kawamura, E. C. LaCasse andJ. Gariepy, Biochemistry, 1998, 37, 5798–5809; A. Pini,C. Falciani and L. Bracci, Curr. Protein Pept. Sci., 2008, 9,468–477.
13 P. I. Kitov and D. R. Bundle, J. Am. Chem. Soc., 2003, 125,16271–16284.
14 B. R. McNaughton, P. C. Gareiss and B. L. Miller, J. Am. Chem.Soc., 2007, 129, 11306–11307.
15 A. Furka, F. Sebestyen, M. Asgedom and G. Dibo, Int. J. Pept.Protein Res., 1991, 37, 487–493; K. S. Lam, S. E. Salmon,E. M. Hersh, V. J. Hruby, W. M. Kazmierski and R. J. Knapp,Nature, 1991, 354, 82–84; T. Lescrinier, C. Hendrix, L. Kerremans,J. Rozenski, A. Link, B. Samyn, A. V. Aerschot, E. Lescrinier,R. Eritja, J. V. Beeumen and P. Herdewijn, Chem.–Eur. J., 1998, 4,425–433.
16 B. B. Brown, D. S. Wagner and H. M. Geysen, Mol. Diversity,1995, 1, 4–12.
17 M. J. Churcher, C. Lamont, F. Hamy, C. Dingwall, S. M. Green,A. D. Lowe, J. G. Butler, M. J. Gait and J. Karn, J. Mol. Biol.,1993, 230, 90–110.
18 T. D. Schneider and R. M. Stephens, Nucleic Acids Res., 1990, 18,6097–6100.
This journal is �c The Royal Society of Chemistry 2009 Mol. BioSyst., 2009, 5, 1070–1073 | 1073